search

robertmartin8 / PyPortfolioOpt

Financial portfolio optimisation in python, including classical efficient frontier, Black-Litterman, Hierarchical Risk Parity
https://pyportfolioopt.readthedocs.io/
MIT License
4.47k stars 951 forks source link

Undocumented shrinkage estimators #41

Closed schneiderfelipe closed 5 years ago

schneiderfelipe commented 5 years ago

Hi, #20 has offered two variants on the Ledoit-Wolf shrinkage estimator that are undocumented. I did some backtests and I believe that the single factor target may be better than the default constant variance target.

I have two questions regarding this:

  1. Is there a reason why the single factor target isn't documented?
  2. Is there a theoretical reason why the single factor might be better?
robertmartin8 commented 5 years ago

Hi Felipe,

I've got a very very basic overview present in the docs, with a link to the original journal article for people to read more.

For your second question, I'm afraid I don't have an answer – and it's not something that Ledoit and Wolf's papers discuss either. It is still very much an empirical topic, so I don't think one can do better than what you've done by backtesting it.

Best, Robert

schneiderfelipe commented 5 years ago

@robertmartin8

Thanks for the answer! I didn't see that part of the docs, sorry!